Nearly half of patients undergoing coronary artery bypass grafting (CABG) nationwide are exposed to red blood cell (RBC) products. While large volumes (3+ units) of transfusions may be delivered to a given patient to preserve life in cases of acute blood loss, evidence suggests that even small amounts (1-2 units) of RBCs often transfused to address poor oxygen delivery are associated with a 16% increase in a patient's risk of mortality and a 27% increase in morbidity, even after case mix adjustment. It is vitally important to allocate RBC products appropriately, given both that cardiac surgery utilizes between 20-25% of the total national blood supply, and the Red Cross reports that 32 of their 36 regions had less than a day's supply available to meet hospital needs. In short, it is important to improve our understanding of how and in what context decisions are made regarding 1-2 units of RBCs, given its association with morbidity, mortality and resource utilization. Growing evidence suggests that some transfusions may be discretionary. In the state of Michigan, nearly 40% of CABG patients are exposed to 1-2 units of RBCs, although the absolute rate varies 32% across institutions. Center-specific variation in transfusion practices is likely attributed to thelack of consensus regarding the indications, setting and hematocrit trigger for transfusions. Given this uncertainty, transfusion rates may be the consequence of differences in organizational (e.g. protocols, types of decision-makers, lack of performance feedback regarding transfusion practice) and provider (e.g. knowledge, beliefs concerning the benefit/harm of transfusions, and inclination to transfuse) factors. Interventions to reduce the rate of unnecessary transfusions first require identifying the set of determinants (at the organizational or provider level) that mot fully explain the observed variation in RBC utilization across regional medical centers. We will use the prospective data and infrastructure of the Michigan Society of Thoracic and Cardiovascular Surgeons Quality Collaborative (MSTCVS-QC), a consortium of all 33 cardiac surgical programs in Michigan, to: (1) Develop, pilot, and implement surveys to hospitals and clinical providers within the state of Michigan to describe determinants of 1-2 units of RBC transfusions during cardiac surgery, and (2) Identify the primary organizational and provider characteristics contributing to variability in transfusions. Results from this study will set the sage for a behavioral modification study aimed at reducing the rate of RBC transfusions in the setting of cardiac surgery.